<![CDATA[Evidence on the pricing of green bonds remains mixed across global markets, particularly in emerging economies where macro-financial risks often overshadow sustainability commitments. Existing research rarely integrates sovereign risk, inflation dynamics, and structural market depth into assessments of whether ESG performance lowers financing costs, leaving the mechanisms behind cross-country variation insufficiently understood. This study identifies the key determinants of green bond yields worldwide and evaluates whether strong sustainability performance effectively reduces borrowing costs, with a specific focus on Indonesia as a representative emerging market. The analysis draws on Signaling Theory, which views ESG commitments as credibility-enhancing disclosures, and on the semi-strong Efficient Market Hypothesis, which suggests that markets incorporate sustainability information only after accounting for fundamental macroeconomic risks. Using 1,362 green bonds issued between 2014 -2023, the study applies a two-layer analytical framework combining Extreme Gradient Boosting with Shapley Additive Explanations to capture non-linear yield dynamics and quantify each variable’s marginal contribution. Robust tests examine stability across pre-crisis, crisis, and post-crisis regimes. Structural and macroeconomic factors especially domicile is the dominant driver of yield formation. ESG attributes remain relevant, but the social pillar exerts the strongest influence, while environmental and governance dimensions function largely as baseline compliance indicators. Indonesia displays a distinctive high-yield, high-ESG pattern driven by inflation pressure, sovereign-risk premia, and shallow market depth. ESG advantages reduce yields only after core macro-financial risks are incorporated. Strengthening macro stability and institutional credibility is essential for sustainability performance to translate into lower financing costs in emerging markets. This study provides one of the first large-scale, cross-country assessments using machine learning and explainable AI to reveal how structural constraints moderate the effect of ESG performance on green bond pricing.]]>
